hpNPCs, Mitochondria quantification using ATP5A staining
Sebastian Schmidt
03 August, 2023
1 Loading of data
1.1 Data set
First, we load, filter, and merge the data sets.
How does the data set looks like
1.2 Set tresholds
Applied tresholds are indicated by grey horizontal line.
1.2.1 Mean_Puncta_mito_AreaShape_Area
1.2.2 Mean_Puncta_mito_Number_Object_Number
1.2.3 mito_MeanArea
1.2.4 mito_MeanCount
1.2.5 mito_MeanLength
1.2.6 Branchpoints
1.3 Counts per sample
#Apply tresholds
data <- subset(data, Mean_Puncta_mito_AreaShape_Area < 200)
data <- subset(data, Mean_Puncta_mito_Number_Object_Number < 1200)
data <- subset(data, mito_MeanArea < 0.04)
data <- subset(data, mito_MeanCount < 0.2)
data <- subset(data, mito_MeanLength < 0.1)
data <- subset(data, Branchpoints < 200)
#Save data set
write.csv(data, file = "results/tables/data_mito.csv")Cell counts per cell line:
#data <- read.csv("results/tables/data_mito.csv")
table(data$Metadata_SampleID)##
## i1JF-R1-018 iG3G-R1-039 i1E4-R1-003 iO3H-R1-005 i82A-R1-002 iJ2C-R1-015
## 110 118 125 118 101 114
## iM89-R1-005 iC99-R1-007 iR66-R1-007 iAY6-R1-003 iPX7-R1-001 i88H-R1-002
## 87 105 89 167 131 76Mean cell count:
mean(table(data$Metadata_SampleID))## [1] 111.752 Visualize mitochondrial parameters
Various mitochondrial parameters are visualized for each patient-derived cell line as well as for the disease state Mean Ctrl levels are indicated by grey horizontal line.
2.1 mito Area
2.1.1 each sample
2.1.2 disease-state
2.2 mito Count
2.2.1 each sample
2.2.2 disease-state
2.3 mito Intensity
2.3.1 each sample
2.3.2 disease-state
2.4 mito Mean Area
2.4.1 each sample
2.4.2 disease-state
2.5 mito Mean Count
2.5.1 each sample
2.5.2 disease-state
2.6 mito Number Branch Ends
2.6.1 each sample
2.6.2 disease-state
2.7 mito Number Branchpoints
2.7.1 each sample
2.7.2 disease-state
2.8 mito Skeleton Length
2.8.1 each sample
2.8.2 disease-state
2.9 mito Mean Skeleton Length
2.9.1 each sample
2.9.2 disease-state
3 Statistical testing using linear mixed effects models
Nested approach (“Mitochondrial Parameter” ~ Disease_state + (1 | Disease_state:Metadata_SampleID)) to compensate for dependencies within the groups.
3.1 mito Area
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula:
## Mean_Puncta_mito_AreaShape_Area ~ Disease_state + (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## 13185.5 13206.3 -6588.7 13177.5 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2591 -0.7390 -0.1397 0.5596 3.6687
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 51.01 7.142
## Residual 1066.69 32.660
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 73.945 3.475 21.279
## Disease_statesPD 6.561 4.563 1.438
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.761## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: Mean_Puncta_mito_AreaShape_Area
## Chisq Df Pr(>Chisq)
## Disease_state 2.0668 1 0.15053.2 mito Count
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula: Mean_Puncta_mito_Number_Object_Number ~ Disease_state + (1 |
## Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## 19004.5 19025.3 -9498.2 18996.5 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8091 -0.8211 -0.1309 0.7812 2.5681
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 3299 57.44
## Residual 81869 286.13
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 565.65 28.35 19.955
## Disease_statesPD -10.11 37.24 -0.272
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.761## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: Mean_Puncta_mito_Number_Object_Number
## Chisq Df Pr(>Chisq)
## Disease_state 0.0738 1 0.78593.3 mito Intensity
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula: Mean_Puncta_mito_Intensity_MeanIntensity_Corr_mito ~ Disease_state +
## (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## -4093.7 -4072.9 2050.8 -4101.7 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.1579 -0.7031 -0.1618 0.6287 4.3208
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 0.0007024 0.02650
## Residual 0.0026666 0.05164
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.14448 0.01205 11.991
## Disease_statesPD -0.01822 0.01579 -1.154
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.763## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: Mean_Puncta_mito_Intensity_MeanIntensity_Corr_mito
## Chisq Df Pr(>Chisq)
## Disease_state 1.3317 1 0.24853.4 mito Mean Area
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula: mito_MeanArea ~ Disease_state + (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## -9414.2 -9393.4 4711.1 -9422.2 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6079 -0.6999 -0.2744 0.4379 3.8921
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 1.68e-06 0.001296
## Residual 5.13e-05 0.007163
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.0102080 0.0006527 15.639
## Disease_statesPD 0.0010092 0.0008580 1.176
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.761## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: mito_MeanArea
## Chisq Df Pr(>Chisq)
## Disease_state 1.3836 1 0.23953.5 mito Mean Count
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula: mito_MeanCount ~ Disease_state + (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## -4425.0 -4404.2 2216.5 -4433.0 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5945 -0.8088 -0.2109 0.6250 2.9401
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 7.809e-05 0.008837
## Residual 2.116e-03 0.046002
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.075443 0.004397 17.158
## Disease_statesPD -0.002436 0.005778 -0.422
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.761## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: mito_MeanCount
## Chisq Df Pr(>Chisq)
## Disease_state 0.1777 1 0.67333.6 mito Number Branch Ends
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula: ObjectSkeleton_NumberBranchEnds_mito_Skeleton ~ Disease_state +
## (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## 10055.6 10076.4 -5023.8 10047.6 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6671 -0.7047 -0.1464 0.5690 4.9810
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 1.512 1.23
## Residual 104.186 10.21
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 16.4336 0.6966 23.592
## Disease_statesPD -0.3257 0.9177 -0.355
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.759## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: ObjectSkeleton_NumberBranchEnds_mito_Skeleton
## Chisq Df Pr(>Chisq)
## Disease_state 0.1259 1 0.72273.7 mito Number Branchpoints
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula: Branchpoints ~ Disease_state + (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## 13178.2 13199.0 -6585.1 13170.2 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6695 -0.7268 -0.1804 0.4718 4.2733
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 27.24 5.219
## Residual 1065.67 32.645
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 45.253 2.705 16.728
## Disease_statesPD 1.675 3.558 0.471
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.760## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: Branchpoints
## Chisq Df Pr(>Chisq)
## Disease_state 0.2217 1 0.63773.8 mito Skeleton Length
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula:
## ObjectSkeleton_TotalObjectSkeletonLength_mito_Skeleton ~ Disease_state +
## (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## 18029.6 18050.4 -9010.8 18021.6 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5013 -0.7576 -0.1928 0.5112 4.5958
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 552 23.49
## Residual 39843 199.61
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 273.809 13.427 20.393
## Disease_statesPD -6.965 17.691 -0.394
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.759## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: ObjectSkeleton_TotalObjectSkeletonLength_mito_Skeleton
## Chisq Df Pr(>Chisq)
## Disease_state 0.155 1 0.69383.9 mito Mean Skeleton Length
## Linear mixed model fit by maximum likelihood ['lmerMod']
## Formula:
## mito_MeanLength ~ Disease_state + (1 | Disease_state:Metadata_SampleID)
## Data: data
##
## AIC BIC logLik deviance df.resid
## -6350.7 -6329.9 3179.4 -6358.7 1337
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.7079 -0.7971 -0.1527 0.6045 2.9565
##
## Random effects:
## Groups Name Variance Std.Dev.
## Disease_state:Metadata_SampleID (Intercept) 1.187e-05 0.003445
## Residual 5.049e-04 0.022471
## Number of obs: 1341, groups: Disease_state:Metadata_SampleID, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.033490 0.001805 18.550
## Disease_statesPD -0.001349 0.002375 -0.568
##
## Correlation of Fixed Effects:
## (Intr)
## Diss_sttsPD -0.760## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: mito_MeanLength
## Chisq Df Pr(>Chisq)
## Disease_state 0.3224 1 0.5702